Tumor-rejection antigen genes have mainly been identified for malignant melanomas, and accordingly, cancer immunotherapies utilizing these genes are developed. Specifically, with recognition of the importance of CD8-positive T cells in anti-tumor immune responses, cancer vaccine therapy that induces tumor-specific CD8-positive T cells in vivo has received attention, and is being applied to various clinical applications. In addition, the mechanism by which peptides consisting of approximately ten amino acid residues activate T cells via the Class I pathway, by the assistance of various costimulatory molecules, to induce tumor-specific cytotoxic T cells (CTLs) was also elucidated. Furthermore, peptides restricted to individual HLA molecules are being actively identified.
However, complete control of tumors is currently impossible. This may be due to tumor cell heterogeneity, a reduction or disappearance of MHC-Class I expression in tumor cells, and the absence of target molecules in the tumor cells. Furthermore, the currently identified tumor antigen peptides exist in some types of tumors, but can not exist in all tumor types. Thus, to resolve these problems, the present inventors did not use tumor cells as target cells, but rather focused on the endothelial cells of tumor vessels. More specifically, endothelial cells have hardly any problems involving a decrease or disappearance of MHC-Class I expression, or heterogeneity. Thus, if CTLs that target tumor vessels can be induced, problems in conventional cancer vaccine therapy, such as the disappearance of Class I and absence of a target molecule, may be overcome, regardless of the type of tumor, and excellent therapeutic effects can be anticipated. Studies on tumor angiogenesis were started from a pioneering hypothesis proposed by Folkman et al. in the 1970's, and have been conducted from various angles. Many studies have been carried out on vascular endothelial growth factor (VEGF)-VEGF receptors (VEGFRs) to evaluate their significance in tumor angiogenesis. Angiogenesis inhibitors have been vigorously developed as target-oriented drugs, particularly in cancer therapy, and are already being clinically tested. However, therapies that use this concept for cancer vaccine treatments are not yet in use. One of the reasons may be immunotolerance to VEGFR, which is expressed in normal cells. However, in the 1990's, Plate, Millauer, and Risau et al. confirmed that VEGFRs were strongly expressed in the endothelial cells of tumoral tissues. In addition, the immune response to autoantigens such as CEA and HER/neu, which are also expressed in normal cells, is not necessarily one of immunotolerance. Thus, the present inventors reasoned that VEGFRs may be used as targets for cancer vaccine therapy.
Recently, it was reported that active immunization against VEGFRs can inhibit angiogenesis in tumors, and metastasis (The Journal of Experimental Medicine, 2002, 195:12, 1575-1584). However, this literature merely used soluble VEGFR proteins, and made no investigation of the amino acid sequences of effective peptides.